Menlo Ventures, a Menlo Park, CA-based venture capital firm, released The State of Generative AI in the Enterprise.
The report is based on a survey of 450+ enterprise executives that details current generative AI adoption and, based on that data, makes projections for where the next wave of AI innovation will happen.
According to the report, enterprise investment in generative AI remains small, totaling $2.5 billion this year–significantly less than the $70 billion invested in traditional AI and the $400 billion spent on cloud software. At the same time, the data shows that the number of enterprises using AI increased 7% in 2023, with AI spend growing by 8% and eating into total enterprise tech spend, which only grew by 5%.
Today, the majority of this spend goes to incumbents. The report explains how “Big Tech” companies and category leaders were able to mobilize and capture a significant portion of the market, leveraging advantages in brand, scale, budget, and engineering resources to integrate AI capabilities into established products. As a result, the market share held by incumbents is currently estimated to be $1.2 billion, leaving little space for challengers to gain traction.
To unlock the enterprise market, AI solutions must upend the status quo by demonstrating significant gains in productivity, replacing old methodologies, and rewriting workflows in ways that feel entirely new and radically different.
The report identifies three areas that offer enormous potential for startups to break out:
- Vertical AI: Research shows that the marketing and legal industries currently lead the way in vertical generative AI adoption. In industry-specific applications, AI will reinvent human-machine collaboration, becoming the driver for end-to-end automation rather than merely a copilot or collaboration platform.
- Horizontal AI: Horizontal solutions are popular because they can be used across industries and departments, increasing workflow efficiency beyond what was previously possible. As AI becomes more capable of reasoning, collaborating, communicating, learning, and predicting, next-generation workflow tools will not only allow machines to augment or automate routine tasks, but—with advanced approaches like agents and multi-step reasoning—take on work that only humans could do before.
- The modern AI stack: New generative capabilities require new tools for building LLM apps, including databases, serving infrastructure, data orchestration, and pipelines. Although still a “work in progress,” the modern AI stack attracts the largest percentage of enterprise AI investment, making it the biggest new market in the generative AI domain and an attractive focus for startups.